We postulated that CCR2-driven activation of the transcription factor NF-κB plays a critical role in dendritic cell (DC) maturation (e.g., migration, costimulation, and IL-12p70 production), necessary for the generation of protective immune responses against the intracellular pathogen Leishmania major. Supporting this notion, we found that CCR2, its ligand CCL2, and NF-κB were required for CCL19 production and adequate Langerhans cell (LC) migration both ex vivo and in vivo. Furthermore, a role for CCR2 in upregulating costimulatory molecules was indicated by the reduced expression of CD80, CD86, and CD40 in Ccr2−/− bone marrow-derived dendritic cells (BMDCs) compared with wild-type (WT) BMDCs. Four lines of evidence suggested that CCR2 plays a critical role in the induction of protective immunity against L. major by regulating IL-12p70 production and migration of DC populations such as LCs. First, compared with WT, Ccr2−/− lymph node cells, splenocytes, BMDCs, and LCs produced lower levels of IL-12p70 following stimulation with LPS/IFN-γ or L. major. Second, a reduced number of LCs carried L. major from the skin to the draining lymph nodes in Ccr2−/− mice compared with WT mice. Third, early treatment with exogenous IL-12 reversed the susceptibility to L. major infection in Ccr2−/− mice. Finally, disruption of IL-12p70 in radioresistant cells, such as LCs, but not in BMDCs resulted in the inability to mount a fully protective immune response in bone marrow chimeric mice. Collectively, our data point to an important role for CCR2-driven activation of NF-κB in the regulation of DC/LC maturation processes that regulate protective immunity against intracellular pathogens.
Dendritic cell (DC) maturation is a highly regulated cellular event that includes at least three processes: migration, upregulation of costimulatory surface molecules (along with Ag presentation), and the release of polarizing cytokines, such as IL-12 (1–4), by Th cells. Models describing events occurring in each of the three processes described above have emerged (5, 6), yet the full array of factors coordinating DC maturation remains unknown. To date, these models build on disjointed evidence suggesting that DC maturation processes are driven by gene programs triggered by transcription factors, such as NF-κB (7–10).
The study of the biology of DCs residing in the skin, such as Langerhans cells (LCs) and dermal DCs, has provided significant insight into the mechanisms underlying DC maturation (11–13), particularly as it relates to migration (5, 6). For instance, LC migration is dependent on selective changes in the levels of chemoattractant molecules, known as chemokines, and to changes in LC responsiveness to these chemoattractants, a consequence in the upregulation of chemokine receptors such as CCR7 (6, 14, 15). CCL19, a major ligand for CCR7, is produced by maturing DCs and by high endothelial venules, creating a gradient from the site of its release to the lymphatics (5, 6, 16, 17). Cognate interactions between this gradient and CCR7, present on the DC’s surface, drives the cells into the draining lymph nodes (DLNs) (6, 17, 18). Moreover, it is likely that induction of CCL19 production and the parallel maturation process are initially triggered by chemokines produced early on, which have the ability to provide initial signals leading to NF-κB–dependent gene transactivation.
Several models have been widely used in the literature to study LCs. In this study, we took advantage of an ex vivo model of ear skin explants and an in vivo model of contact hypersensitivity. (19–21). In the former model, skin from mouse ears is separated into two layers and placed in a petri dish containing culture media. The physical separation of the two skin layers induces LCs migration from the epidermis into the dermal lymphatics. LCs are collected as they fall off from the lymphatic channels directly into the petri dish (19). The in vitro model involves painting the skin with FITC plus an irritant to induce the migration of LC carrying FITC to the DLNs (20, 21). In a previous study, our laboratory successfully tested the hypothesis that, in addition to CCR7, CCR2 plays a key role in LC trafficking during inflammation using these models.
We showed that mice with genetic inactivation of CCR2 exhibited impaired LC migration and an inability to control infection with the intracellular pathogen Leishmania major (21). Notably, our group (and others) has previously shown that CCR2-dependent signals result in the induction of NF-κB translocation in astrocytes, but it remains unknown whether CCR2-dependent activation of NF-κB is also linked to CCR2-dependent modulation of DC function. This seems plausible considering that DC maturation has been shown associated with changes in the expression of CCR2 (22) and that maturation is thought to be modulated by NF-κB, as described above.
We surmised that migration defects in CCR2-deficient mice might be a component of the broader disturbances affecting DC maturation. Hence, using mice with genetic inactivation of CCR2, we sought to investigate the impact of CCR2 in DC migration and maturation, aiming to test the hypothesis that CCR2-driven activation of NF-κB could play a critical role in DC maturation processes, which in turn might be necessary for the generation of protective immune responses against the intracellular pathogen L. major. In this paper, we show that CCR2, potentially through the induction of NF-κB translocation, plays a role in DC maturation involving CCL19 production, LC migration; Ag presentation with upregulation of costimulatory molecules in DCs; and the release of the Th1-inducing cytokine IL-12. Furthermore, we also provide evidence suggesting that IL-12p70, production by radioresistant (RR) cells, such as LCs, contributes to the generation of protection against L. major infection. Thus, the role of CCR2 in the elicitation of protective immunity is 2-fold; on one hand, CCR2 modulates the amount of IL-12 produced by DCs, and on the other hand, CCR2 regulates the migrating number of IL-12p70 producing LCs.
Materials and Methods
4 (LTC4) was from Cayman Chemical (Ann Arbor, MI), and recombinant proteins CCL19 (ED50, 3–15 ng/ml; endotoxin < 0.1 ng/μg) and CXCL1 (ED50
Mice, CCL19 blockade, and bone marrow transplantation
The Institutional Animal Care and Use Committee approved all protocols. The genetic inactivation of CCR2 mice and genotyping has been described previously (23). Wild-type (WT) and gene knockout (KO) mice were on the C57BL/6J background (n > 10 generations). In some experiments, we used WT and CCR2 KO mice of C57BL/6 × 129 background, and we obtained similar results to those in the C57BL/6J. Mice were born and bred under specific pathogen-free conditions at the University of Texas Health Science Center (San Antonio, TX). SCID mice were purchased from Taconic Farms (Germantown, NY). C57BL/6J IL-12p35−/− mice were purchased from The Jackson Laboratory (Bar Harbor, ME).
CCL19 Ab neutralization was performed in WT C57BL/6J, as described previously (16L. major infection and weekly thereafter. Ear thickness was recorded on a weekly basis, mice were sacrificed 4 wk postinfection, and parasite burdens were determined in DLNs.
For bone marrow (BM) transplantation, recipient mice received 1000 cGy (lethal dose) irradiation from a 137Cs source 6–8 h before the cell transfer. Donor mice were euthanized, and BM was harvested, as described below. BM cell suspensions were depleted of RBCs using RBC lysis buffer, resuspended in PBS, and transferred into recipient mice (10 × 106 in 200 μl PBS) via tail vein injection. Recipient mice were infected 8 wk after BM transfer.
Parasites, infection, and tracking of infected LCs
L. major clone VI (MHOM/IL/80/Friedlin) was a gift from Dr. D.L. Sacks (Laboratory of Parasitic Diseases, National Institutes of Health, Bethesda, MD). The virulence of the L. major strain was maintained by propagating it in mice (21). Protocols similar to those used to prepare soluble Leishmania donovani Ags were used to make soluble L. major Ag (21). The intradermal route was used to infect both ears with 2 × 106 or 1 × 107 stationary-phase L. major promastigotes resuspended in 20 μl PBS (21). Higher doses of parasites were used with similar results in short-term experiments. A dial-thickness gauge caliper (Fred V. Fowler, Newton, MA) was used to measure ear thicknesses at weekly intervals for 5 wk (21). The parasites were grown in medium 199 containing 20% FBS and supplemented with hemin, adenine, and antibiotics. The limiting dilution culture method was used to quantify parasite burden (21). In brief, relevant organs were weighed and homogenized between two frosted slides. The homogenate was placed in 96-well plates in triplicate, and 5-fold serial dilutions were performed in complete medium 199. Parasites were cultured for 2 wk at 27°C. The plates were scored for the presence of motile parasites in individual wells using an inverted microscope at ×200 magnification. The parasite scores are the log values of the reciprocal from the last positive dilution.
PKH26 (Sigma-Aldrich) labeling of L. major was performed as described by Quinones et al. (24). A minimum of 2.5 × 107 PKH-labeled parasites were used to infect mice. Injection with a lower amount of parasites made the identification of cells carrying the parasite unreliable. This adjustment in the dose was performed to account for the loss of parasite viability associated with PKH labeling and to increase our ability to detect fluorescent-labeled parasites in vivo. Parasite viability was preserved following PKH labeling and in vivo inoculation. Confirming the pathogenicity and viability of labeled parasites in vivo, we found that infection with PKH-labeled parasites did in fact result in chronic infection as documented by parasite recovery from the DLNs of infected animals. Three days following intradermal injection of PKH-labeled L. major, mice were euthanized, and FACS was used to determine the proportion of CD11c+ cells carrying PKH-labeled L. major in the DLNs. Cell sorting and subsequent culture of PKH+CD11c+ cells under conditions favoring parasite growth gave rise to L. major, indicating that PKH labeling of L. major did not significantly affect viability and pathogenicity (data not shown). We inferred that LCs constituted a large proportion of the population of PKH+CD11c+-labeled cells considering that the majority of these cells also expressed CD11b+ and DEC205+ and were CD8α−/low (data not shown). Furthermore, in vitro, PKH+CD11c+ cells promoted Ag-specific proliferation of purified T cells derived from the L. major-infected mice.
LC migration to lymph nodes and NF-κB in vivo inhibition
The protocols described by Sato et al. (21) were used to track the migration of FITC+ LCs, from the ear skin to the DLNs in uninfected animals. In brief, freshly prepared 5% FITC in 1:1 acetone/dibutylphthalate was painted on both sides of the ears (10 μl each) and on the abdomen (400 μl) (to quantify the number of FITC-positive DLNs following Flt3 ligand [Flt3L] treatment). The DLNs were harvested 48 h later and analyzed by FACS or frozen in OCT compound (Sakura Finetek, Torrance, CA), and 5-μm cryostat sections were prepared and examined by fluorescence microscopy (BX60 Olympus microscope; Olympus, Melville, NY). Animals were given a 0.1-ml i.p. and 10-μl intradermal (ear) injection of the NF-κB inhibitors pyrrolidine dithiocarbamate (PDTC; 6 μg/μl) and MG-132 (5 μM) 1 h before FITC painting or ear explant experiments.
Cell migration from ear skin explants
As previously described (19, 21), ears were dissected, rinsed in 70% ethanol with vigorous shaking, and allowed to dry for 15 min. The ventral and dorsal sheets of ear skin were separated with a pair of fine forceps. The two leaflets were transferred dermal side down for 24 or 72 h in a 10-cm petri dish containing culture media. The loosely adherent populations of cells that spontaneously emigrated out of the dermal layers were recovered by incubating the dermal layers in PBS containing 2 mg/ml glucose for 20 min at 37°C. The collected cells were pooled together, filtered through a 70-μm nylon cell strainer, washed with PBS, and then stained. In these experiments, LCs were identified as I-Ab bright and DEC-205+. In selected experiments, different compounds were added to the culture media of the ear explants.
Generation BM-derived DC and in vitro infections
BM-derived DCs (BMDCs) were generated as described previously (25). Briefly, BM was collected from 6- to 8-wk-old mice and cultured in the presence of: recombinant murine (rm)GM-CSF (50 ng/ml) and rmIL-4 (1 ng/ml). BMDC cultures have been shown to contain a large proportion of mature DCs (25). For infection studies, BMDCs were cultured with live L. major parasites in a 1:10 ratio.
In vitro Ag presentation activity
The Ag-specific syngeneic proliferation assays adapted for this experiment have been described previously (21Ccr2−/− mice. After 48 h, culture supernatants were harvested, and cytokine levels were measured by ELISA.
In a set of experiments, 10 d prior to infection, WT and Ccr2−/− mice received a daily i.p. dose of 10 μg recombinant human Flt3L, as previously described (26, 27), and followed during a period of 7 wk. Groups of uninfected animals were used for corroboration of the Flt3L effect on DCs by FACS and analysis of DC migration using ear explants and FITC painting.
Administration of rmIL-12 or LPS and measurement of endogenous heterodimeric IL-12p70 production
Mice were injected i.p. with 0.5 μg rmIL-12 (ED50 < 0.1 ng/ml, endotoxin < 0.1 ng/μg; PeproTech, Rocky Hill, NJ) or 100 μg LPS at the time of infection and repeated at 24 and 48 h postinfection. The IL-12 administration was a modification from a protocol described by Heinzel et al. (28). The administration period of IL-12 was shortened from 7 to 2 d given that we wanted to capture the early release of this cytokine. A total of 1 μg/ml LPS and 10 ng/ml IFN-γ (ED50 0.1 ng/ml, endotoxin < 0.1 ng/μg; PeproTech) were used to induce IL-12p70 production in DLNs using protocols described previously (24). At different time points (6, 12, 24, and 48 h poststimulation with LPS plus IFN-γ), cultured supernatants were collected. ELISA was used to determine IL-12p70 levels in the supernatants, as described previously (24). Unless indicated otherwise, background values obtained in the absence of stimulation are subtracted from the data presented.
NF-κB translocation in ear explants
Ears from WT and Ccr2−/− mice were split and immediately frozen (0 h) or left in culture for 2 h and then frozen. Subsequently, ears were cut into small pieces and homogenized, and the nuclear proteins were extracted using the nuclear protein extraction kit from Panomics (Fremont, CA) following the manufacturer’s protocols. NF-κB was quantified in the nuclear extracts using a previously validated beads-based system from Panomics. The coefficient of variation for the assay was under 5% based on the company’s data. The amount of NF-κB minus background signal was normalized based on 1) the levels of the housekeeping gene transcription factor IID and 2) by the total amount of nuclear extracts originally recovered from the ear explants.
RNase protection assay
RNase protection assay (RPA) was performed as described previously (23, 29). Total RNA was isolated from homogenized ear explants and used to perform the RPA using RiboQuant Multiprobe RNase Protection Assay Kit (BD Biosciences). The data are presented as the ratio of the densitometric signals of a housekeeping gene L32 to the mRNA for the gene of interest.
CCL19 protein levels from homogenized ear explants at 6, 12, 24, and 72 h were analyzed using a custom ELISA. Protein tissue lysates were prepared, as described previously (24
Ag-specific cytokine production in infected mice
Single-cell suspensions (5 × 106/ml) were cultured with or without 50 μg/ml soluble L. major Ag in 24-well plates, as described previously (21). After 48 h of culture, supernatants were harvested, and ELISA was used to determine the IL-4, IL-5, and IFN-γ protein levels. The range of detection by ELISAs for IL-4 was 2.0–125 and 15.6–1000 pg/ml for IL-5 and IFN-γ.
Statistical analysis and data modeling
Data represent the mean ± SD. Groups were analyzed with Stata (StataCorp, College Station, TX) or SPSS (Chicago, IL) statistical software. According to the number of groups and the distribution (normally distributed or not), nonpaired t test, one-way ANOVA, Kruskal-Wallis, Mann-Whitney U, or Fisher’s exact test were performed. Statistical significance was accepted at p < 0.05. Pathways Analysis Software (Ingenuity Systems, Redwood City, CA) was used to visualize potential pathways suggested from our data and emerging from published literature by looking at interactions among CCR2, CCL2, CCL19, NF-κB, and IL-12.
Impaired LC migration in CCR2 KO mice is associated with reduced CCL19 production
We have shown that compared with WT mice, Ccr2−/− mice exhibit decreased numbers of LCs migrating out of ear skin explants (21). This reduced migration is seen despite comparable numbers of resident LCs in the epidermis of WT and Ccr2−/− mice (21). A similar outcome has been described in mice lacking CCR7 or its ligand CCL19, suggesting that the CCR7–CCL19 axis is essential in coordinating LC migration (6, 16, 17, 30). Thus, we hypothesized that there could be a connection between signals generated by CCR2 and the activation of the CCR7–CCL19 axis that occurs during DC maturation. To test this notion, we first measured the levels of expression of CCR7 and CCL19 mRNA in 24- and 72-h ear skin explants from WT and Ccr2−/− mice.
Compared with the WT mice, ear skin explants derived from Ccr2−/− mice significantly expressed lower levels of CCL19 but not CCR7 or CCL2, a control chemokine (Fig. 1A). Furthermore, this reduction in mRNA translated into significantly lower CCL19 protein concentrations in the ear skin explants of Ccr2−/− compared with WT mice (Fig. 1B). Because DCs are a major source of CCL19 in the ear explants (16), it seemed plausible that Ccr2-null DCs had a reduced ability to produce this chemokine.
Next, we surmised that if LC migration in Ccr2−/− mice was defective, because of lower levels of CCL19 in the explants, then reconstitution of CCL19 to normal levels should ameliorate the LC migration defect. To determine whether this was indeed the case, we added CCL19 into the culture medium of the ear skin explants of WT and Ccr2−/− mice and analyzed LC migration. We found that the addition of CCL19, but not the unrelated chemokine CXCL1, into the culture medium significantly increased the number of LCs that migrated out from ear explants of WT mice as well as those of Ccr2−/− mice (Fig. 1C). However, the magnitude of the increase induced by CCL19 was comparatively higher in ear explants derived from Ccr2−/− mice (Fig. 1C). Next, we asked whether this increase in LC migration by CCL19 was dependent on LTC4. The rationale for this line of inquiry was provided by two studies that showed that LTC4 is required to induce chemotaxis toward CCL19 in LCs (16) and that CCR2 activation induces the release of LTC4 in a model of airway hyperreactivity (31). We found that addition of LTC4 did not enhance the migration of LC from ear skin explants derived from Ccr2−/− mice (Fig. 1C). Taken together, these results suggest that the LC migration defect seen in Ccr2−/− mice may be due to reduced local upregulation of CCL19.
Possible role of CCL2, the principal ligand of CCR2, and the induction of NF-κB–dependent gene transactivation of DC migration
We then focused on determining the possible upstream and downstream mechanistic pathways that could account for the defect in CCL19 production and the related LC migration defect in Ccr2−/− mice. We envisioned a scenario in which CCL2 released locally acts on its receptor CCR2 to induce the translocation of the transcription factor NF-κB into the nucleus. This transcription factor promotes DC migration manifested by the release of CCL19 and LC migration from the epidermis toward the DLNs. The following converging lines of evidence supported this line of reasoning.
First, we found that, similar to the findings in Ccr2−/− mice, genetic inactivation of the CCR2 ligand CCL2 was also associated with abnormal LC migration (Fig. 1D). As expected, decreased LC migration was accompanied by a reduction in CCL19 mRNA (data not shown). Furthermore, providing a more direct link between the CCL2–CCR2 axis and CCL19, we found that the addition of CCL2 to murine splenocytes induced CCL19 production (Fig. 1E), which was localized in DCs (data not shown). Second, 2 h after ear peeling there was a significant increase in nuclear NF-κB translocation in WT explants (1.85-fold increase from time 0 to 2 h; p = 0.009; n = 12 explants for each time point). However, in Ccr2−/− explants, NF-κB translocation was blunted, failing to reach statistical significance (1.47-fold increase from time 0 to 2 h; p= NS; n = 12 explants for each time point). Third, in vivo systemic treatment of mice with NF-κB inhibitors disrupted LC migration. We used two different compounds routinely used for in vitro and in vivo studies (32–39). Both PDTC (32, 33) and MG-132 (40), administered to mice prior to epidermal FITC painting, inhibited LC migration from ear skin to DLNs (Fig. 1F). A similar effect was seen when inhibition of NF-κB was achieved by adding the inhibitors to cultured ear explants (data not shown).
Finally, we observed that chemical inhibition of NF-κB resulted in a reduction in CCL19 levels in the tissue (Fig. 1G) linking NF-κB and CCL19 and falling in line with our findings that LC migration in Ccr2−/− and Ccl2−/− mice is disrupted. We used two different inhibitors because in general chemical inhibitors are not 100% specific. Therefore, PDTC and MG-142 could have other effects beyond NF-κB inhibition that could also affect LC migration.
Abnormal upregulation of costimulatory molecules in Ccr2−/− DCs
The results presented so far suggest that CCL2-CCR2–mediated induction of NF-κB gene transactivation could be required for CCL19-driven LC migration from the skin to the DLNs. Next, we tested the notion that absence of Ccr2 and the related impairment of DC maturation could also be manifested by a failure to upregulate costimulatory molecules and MHC class II. To test this notion, we studied DC differentiation and maturation occurring in vitro from BM progenitors induced by GM-CSF and IL-4 (24, 25). In this well-characterized in vitro model system, we found that despite comparable numbers of CD11c+ DCs in cultures from WT and Ccr2−/− mice (Fig. 2A), the percentages of DCs expressing costimulatory molecules and MHC class II were significantly lower in the KO mice (Fig. 2B–E). Overall, this difference was also evident when looking at the mean fluorescence intensity values (mean fluorescence intensities in a representative experiment WT versus Ccr2−/− BMDCs included: CD80, 469 versus 384; CD86, 264 versus 289; and I-Ab, 749 versus 443). This finding is in line with the view that CCR2 participates in DC maturation in a broader context.
Susceptibility of Ccr2−/− mice to L. major infection is related to abnormal maturation of DC populations including LCs
We postulated that genetic inactivation of CCR2 leads to susceptibility to L. major infection by disrupting LCs and other DC maturation, such as migration and IL-12p70 production. To test this notion, we first asked whether CCR2-dependent signals influence the migration of LCs carrying L. major from the skin to the DLNs. To this end, we quantified the proportion of DCs carrying PKH-labeled L. major from the epidermis into the DLNs of WT and Ccr2−/− 3 d after intradermal injection. We found that compared with WT mice, Ccr2−/− DLNs contained a smaller proportion of cells likely to constitute PKH-labeled LCs (Fig. 3A).
Next, we determined whether genetic inactivation of CCR2 affected the production of IL-12p70 by DCs. We first measured IL-12p70 production by DLN cells and splenocytes in vitro, because these lymphoid organs contain diverse populations of DCs, which have been shown to be an important source of this cytokine (42–44). This observation was confirmed using intracellular staining for IL-12, along with the surface staining for the DC marker CD11c (data not shown). In this in vitro system, we found that LPS/IFN-γ–induced production of IL-12p70 by splenocytes and lymph node (LN) cells from Ccr2−/− mice was significantly lower than in WT mice (Fig. 3B, 3C). Next, we sought to determine whether CCR2 also regulates IL-12p70 production and other DC maturation processes triggered by exposure L. major. Indeed, using BMDCs as a model system, we found that L. major-induced upregulation of class II expression was reduced in Ccr2−/− DCs compared with WT DCs (Fig. 3D). Likewise, Ccr2−/− BMDCs produced lower levels of IL-12p70 than WT DCs in response to activation with live L. major (Fig. 3E). The reduction in IL-12p70 production seemed to be functionally relevant given that, compared with WT BMDCs, Ccr2−/− BMDCs had a reduced capacity to induce Ag-specific production of IFN-γ by KLH-primed WT T cells (Fig. 3F). Furthermore, there was a correlation between the defect in IL-12 production of DC Ccr2−/− identified in vitro and in vivo. This was suggested by the observation that there was a reduced proportion of mature LCs producing IL-12 (CD205+I-Ab+IL-12+ cells) in the DLNs of L. major-infected Ccr2−/− mice compared with infected WT mice (Fig. 3G).
We surmised that if the defective production of IL-12p70 in Ccr2−/− DCs seen in vitro was an important determinant of susceptibility to L. major infection, then early exogenous administration of this cytokine to Ccr2−/− mice, at the time of exposure to L. major, would switch the susceptible phenotype to resistance. Supporting this posit, administration of rIL-12 to Ccr2−/− mice during the initial 3 d postinfection corrected their inability to mount a protective response to L. major infection (Fig. 4A, 4B). Compared with untreated Ccr2−/− mice, rIL-12–treated Ccr2−/− mice had reduced ear swelling and reduced spleen parasite burdens at 6 wk postinfection (Fig. 4A, 4B). Furthermore, Ag-specific immune responses in IL-12–treated Ccr2−/− mice were characterized by significantly higher Ag-specific IFN-γ and lower IL-4 and IL-5 compared with untreated Ccr2−/− mice (Fig. 4C–E).
An important issue to address was the specificity of exogenous IL-12 administration for both the phenotype of Ccr2−/− mice and the treatment itself. First, the observation that administration of rIL-12p70 did not have any measurable effect in WT mice (Fig. 4A, 4B) rules out a simple overall beneficial effect of IL-12p70 administration in resistant mice strains. Second, the effect of exogenous IL-12p70 administration in the phenotype of Ccr2−/− mice was not a consequence of nonspecific inflammatory effects as suggested by the lack of efficacy of treatment with LPS. In this experiment, mice received three doses of LPS or saline prior to infection with L. major. Doses of LPS in this range are known to promote DC migration and IL-12 production in vivo (45). Ccr2−/− mice were equally susceptible to infection regardless of receiving either PBS or LPS (Fig. 4F). LPS administration was sufficient to induce systemic effects in WT mice, including DC mobilization from the marginal zone to T cell zone in the spleen (data not shown). This outcome was also in line with previous research demonstrating the inability of LPS administration to change a L. major-susceptible phenotype in mice in vivo (46).
Potential role of RR/LC-derived IL-12 in the generation of protective immunity against L. major infection
DC subsets vary in their ability to produce IL-12p70, and consequently, they differ in their capability to polarize T cell responses toward Th1. However, considering all the redundancy in the DC network, it is unclear whether the production of IL-12p70 by any DC subtype is indispensable for induction of protective immune responses. To infer that L. major susceptibility in Ccr2−/− mice is linked to a decreased amount of IL-12 released by LCs, we first needed to demonstrate that IL-12p70 production by LCs does play a nonredundant role in the generation of protective immune responses against L. major. Aiming to provide evidence for a role of LC-derived IL-12, we engineered BM chimeric mice in which IL-12p70 production was deficient in either the BM-derived cell populations, the RR cell compartment, or in both compartments or was normal.
The rationale for the use of BM chimeras mice was the knowledge that a large proportion of LCs is derived from RR/LC progenitors (47) and that therefore selective replacement of all BMDC populations but not LCs could be accomplished in a lethally irradiated and syngeneically reconstituted host. Using WT and IL-12p35−/− mice as donor or recipients, the following four groups of chimeric mice were created. The first group was lethally irradiated IL-12p35−/− recipients of WT mice BM (WT→IL-12p35−/−). In these mice, LCs should be unable to produce IL-12p70; however, all other populations of BMDCs would release normal IL-12p70 levels. Group 2 consisted of lethally irradiated WT recipient mice of IL-12p35−/− BM (IL-12p35−/−→WT). In these mice, normal release of IL-12p70 is expected in LCs but disrupted in all BMDC populations. The third group consisted of IL-12p35−/−recipients of IL-12p35−/− BM, and the fourth group comprised WT recipients of WT BM. The final two groups served as controls because IL-12 production was blunted in both LCs and BMDCs (IL-12p35−/−→IL-12p35−/−), or it was normal in both cell compartments (WT→WT).
After allowing sufficient time for their reconstitution, chimeric mice were infected intradermally with L. major and followed over 7 wk. Ear swelling and DLN parasite burdens were used as indicators of protective immunity. Importantly, we found that the proportion of LCs (CD11c+CD207+ or CD11c+DEC205+ cells) residing in the DLNs 7 wk postinfection was comparable across all four groups of chimeras, indicating that genetic inactivation of IL-12p35−/− was not influencing DC migration (data not shown).
As anticipated, the control groups WT→WT and IL-12p35−/−→IL-12p35−/− were completely resistant or susceptible to L. major infection, respectively (Fig. 4G, Table I). Interestingly, WT→IL-12p35−/− chimeras had significantly more ear swelling and higher DLN parasite burdens than WT→WT mice but less than IL-12p35−/−→IL-12p35−/−, suggesting that IL-12p70 derived from RR cells, such as LCs, was necessary to elicit some degree of protective immunity against L. major (Fig. 4G, Table I). Moreover, IL-12p35−/−→WT and IL-12p35−/−→IL-12p35−/− chimeras were equally susceptible to L. major infection, suggesting that IL-12p70 produced by RR/LCs was insufficient for the elicitation of adequate immune responses against L. major (Fig. 4G, Table I), and therefore, BM-derived populations were an indispensable source of IL-12p70. Taken together, the results from the BM chimeras suggest that IL-12p70 derived from RR cells, namely LCs, contributes but is not sufficient for the elicitation of full protection against intradermal L. major infection.
On the basis of the results above, we surmised that if direct effects of CCR2 on T cells could be ruled out, then at least two interrelated APC factors might underlie the susceptibility in Ccr2−/− mice to L. major infection. The first factor being a reduced amount of IL-12p70 produced by Ccr2−/− DC populations including LCs (Figs. 3A, 3B, 3E, 3F, 4G, Table I). The second, a reduced number of LCs in the DLN, would factor the production of lower amounts of IL-12p70 (Fig. 3C, 3G) and reduced support adequate T cell activation. To address these factors, the following sets of experiments were conducted.
To rule out the possibility that intrinsic defects in T cells were responsible for abnormalities in Ccr2−/− mice, we conducted adoptive T cell transfer studies from WT or Ccr2-null mice into SCID mice. As shown in Fig. 4H and 4I, T cells from WT or Ccr2−/− mice were comparable in their ability to protect SCID mice against the infectious challenge. This finding is in line with our previous report that WT and Ccr2−/− T cells are comparable in their ability to polarize into Th1/Th2 cells (48).
Next, we attempted to gather evidence for a nonredundant role of the reduced number of LCs arriving in the DLNs in Ccr2−/− mice’s susceptibility to L. major infection. We successfully increased only the number of DCs migrating in Ccr2−/− mice without changing the functional maturation defect (e.g., ability to produce IL-12p70; data not shown) using 10-d treatment with Flt3L (Table II). Administration of Ftl3L is a well characterized approach to expand the multiple subsets of DCs, including LCs and dermal DCs (49). Flt3L treatment partially increased the detection of FITC+ LCs in the DLNs of Ccr2−/− mice (Table II) and LC migrating cells out of the ear explants (data not shown). However, this increase was not sufficient to change the susceptibly state of Ccr2−/− mice to L. major infection as PBS-treated and Flt3L-treated Ccr2−/− mice had a similar outcome following L. major infection (Table II, data not shown).
A possible interpretation of this finding is that although the numerical DC/LC defect may be partially corrected by administering Flt3L to Ccr2−/− mice, increasing the availability of DCs seems ineffective because their functional defects (abnormal IL-12p70 release and expression of costimulatory molecules) remained unchanged.
In this study, we show that CCR2-dependent signals triggered by binding of its ligand CCL2 and subsequent nuclear translocation of NF-κB are required for DC maturation and LC migration from the epidermis to the DLNs. Interestingly, CCR2-dependent signals contribute to the production of CCL19, upregulation of costimulatory molecules, and the production of IL-12 (a proposed model is shown in Fig. 5). In particular, the later process was found to be critical for the generation of protective immune responses against L. major infection in resistant mice. Taken together, these data extend our previous reports that the susceptibility of Ccr2−/− mice to L. major infection could, in fact, be secondary to the inability of LCs to produce IL-12 and a failure to mount a protective Th1 response.
The role of the chemokine system in DC biology is classically seen as being critical for cell migration. However, increasing evidence suggests that in addition to their role in chemotaxis, chemokines and their receptors also regulate DC maturation (50). For instance, human DCs treated with neutralizing Abs against the chemokine receptors CCR1 or CCR3 reduced their ability to activate T cells in vitro (51). In mice, CCL19 was found to induce maturation of DCs, resulting in upregulation of costimulatory molecules, production of proinflammatory cytokines, and enhancement of DCs’ ability to activate T cells (52). More recently, evidence consistent with our findings suggests that CCR2-dependent signals are critical for DC maturation (22). The authors showed that genetic inactivation of CCR2 in murine DCs lead to reduced expression of costimulatory molecules upon activation with LPS, reduced allostimulatory capacity, and abnormal migration (22). Unfortunately, the mechanisms underlying these observations were not reported. In this paper, we extend those observations by revealing the potential role of the CCL2, CCR2’s main ligand CCL2, and of NF-κB–mediated signals, as well as the critical role of CCR2 in regulation of IL-12 production by maturating DCs. We also present data supporting the in vivo relevance of these observations in the context of a specific DC subset, namely LCs, and test the relevance of our findings in an in vivo infectious model.
A role for the CCL2–CCR2 axis in LC migration/maturation is also supported by three recent reports. First, human CCL2 transgenic mice exhibit an acceleration of migration of LC from the epidermis into the DLN, following sensitization with haptens, an effect that is accompanied with increased expression of class II and costimulatory molecules (53). Second, CCL2 has been shown to have potent chemoattractant activity for murine LCs in vitro (54). And finally immature DCs respond to CCL2 (55).
Another line of convergent evidence supporting a role for NF-κB translocation by the CCR2 ligand CCL2 is highlighted in recent work from our laboratory (29). Using an EMSA-based approach, we found CCL2-induced nuclear translocation of NF-κB in astrocytes. The effect of CCL2 on NF-κB translocation was 1) time dependent, peaking at 120 min after the addition of CCL2; 2) dose dependent with maximal effects at 100 ng/ml for the dose-range tested; and 3) dependent predominantly on the p65 component of NF-κB (29). We also demonstrated that CCL2-induced NF-κB translocation was markedly diminished in the absence of CCR2. (29).
A logical extension of this work was to investigate the functional implications of CCL2-CCR2 activation of NF-κB in DC biology.
NF-κB is highly expressed by DCs, and it mediates key DC functions (6). Indeed, in line with our findings, previous reports have highlighted that NF-κB strongly enhances DC longevity (56); well-known DC-activating mediators such CD40L promote maturation via sustained activation of NF-κB (57); and inhibition of NF-κB activation blocks maturation of DCs, in terms of upregulation of MHC and costimulatory molecules (9, 58).
The contribution of NF-κB to DC migration has been less studied ex vivo and in vivo. We were able to document blunting in Ccr2−/− skin explants of the rapid upregulation of NF-κB that occurred following ear peeling in WT explants. Highlighting further the importance of this transcription factor, we found that administration of the NF-κB chemical antagonists PDTC and MG-132 to mice significantly reduced LC migration in the FITC-painted model. Certainly, a major concern with the use of the chemical antagonists is the lack of specificity. However, our finding of reduced LC migration in mice receiving PDTC or MG-132 is further supported by reports indicating that the proteasome inhibitor-based (N-benzyloxycarbonyl-Ile-Glu[O-tert-butyl]-Ala-leucinal [PSI]) or adenoviral- mediated antagonism of NF-κB blocks DC maturation events, such as upregulation of costimulatory molecules and production of cytokines (59).
Moreover, very similar to what we show in the current report, NF-κB blockade using decoy oligodeoxynucleotides in vivo was shown to reduce LC migration and impair T cell responses (60). Nonetheless, our findings extend these observations by revealing that blockade of NF-κB is linked to reduced production of CCL19, a critical chemokine for LC migration (16), and further highlight the critical role of the CCL2–CCR2 axis in the induction of NF-κB. Moreover, the data provide a mechanistic backdrop to the observation that chemical antagonism of CCR2 in vivo reduces the magnitude of delayed-type hypersensitivity responses (61), an immunological process that is highly dependent on DC migration from the skin to the DLNs (2, 60). Thus, collectively our data suggest that CCL2-dependent signals acting via CCR2 lead to NF-κB translocation to the nucleus and transactivation of the CCL19 gene (Fig. 5). This possibility is further supported by work documenting that both human and mouse CCL19 promoters have several functional NF-κB binding sites (8, 41).
Our data also indicate that many of the phenotypic consequences of CCR2 inactivation may be due to deficits in NF-κB activation. Indeed, in IL-12p70 production in DCs evidently occurs in response to signals that require NF-κB activation (7, 62, 63).
The observation that inactivation of CCR2 is associated with abnormal production of IL-12p70 by DCs in vitro supports previous reports documenting abnormal IL-12p70 production in Ccr2−/− mice (64), as well as the fact that chemokines are increasingly recognized as regulators of the production of this cytokine by DCs (65). By contrast, in this study, we show the relevance of reduced production of IL-12p70 in Ccr2−/− mice, in vivo, as highlighted by the demonstration that administration of IL-12p70 prior to infection corrects the L. major-susceptible phenotype but does not affect the response of WT mice.
The observation that treatment with LPS prior to infection, a stimulus known to promote DC maturation and mobilization (66), was not able to induce resistance in Ccr2−/− mice suggests that in the absence of CCR2, DCs are incapable of overcoming using the nonspecific stimuli. Thus, effective signals via CCR2 might be indispensable. Aside from CCL2, other CCR2 ligands could provide these signals as suggested by our observation that CCL2-null mice are resistant to L. major infection (67) and that CCL2 signals are not critical for IL-12p70 production in vivo (64).
Our experiments using lethal radiation and BM transplantation suggests that IL-12p70 released by a cell type with RR progenitors, such as LCs contributes, but is not sufficient, to induce protective immunity against L. major. This finding allows us to infer that the susceptible phenotype in Ccr2−/− mice could be related, in part, to abnormalities in IL-12p70 production by LCs and other DC populations. Importantly, the contribution of CCR2 to the generation of protective immune responses goes beyond its role in LC migration and involves the modulation of maturation processes. Indeed, solely increasing the numbers of DCs available in Ccr2−/− prior to L. major infection using Flt3L failed to change their susceptibility. Our work confirms and further extends the role of DCs (68), specifically LCs, in the generation of immune responses against L. major, including the observation that LCs produce IL-12p70 when exposed to L. major parasites (43, 69) and that IL-12p70 released in the early stages of infection is indispensable for the induction of protective immune responses against L. major (70). It has also been shown that production of IL-12p70 by mature DCs is required to generate protective immune responses against L. major, and work by Wiethe et al. (71) showed that immature DCs induce Th2 polarization and are susceptibility to L. major infection as is likely to be the case in Ccr2−/− mice.
Although the relevance of LCs for the generation of protective immune responses against L. major seems to be well established, some reports challenge this notion (72). For instance, skin-derived DCs, including LCs and dermal DCs, were found to migrate poorly to LNs after L. major infection and have been shown to play a minor role in early T cell activation (73). Moreover, resident DCs in LNs (rather than LCs) were responsible for the induction of protective immune responses (74). Of note, in these reports, the infectious challenge was administered via the s.c. route. By contrast, our study used the intradermal route. Clearly, the use of the intradermal route more likely engaged LCs and led to differential immune responses (67, 75). Furthermore, the route of administration is critical given that LCs promote T cell responses to skin Ag, but only under defined conditions (76) and also through our work indicating that the effects of the CCL2–CCR2 axis are modulated by the route of infectious challenge (67). The doses of parasite used also can be a critical factor. The doses of L. major used in this study were significantly higher than expected during human infection and did not use a live vector (sandfly) to inoculate the parasites; therefore, the underlying cellular and molecular physiological process observed may not represent normal responses against infection in a natural setting.
We thank N. Sato, Jason Schaeffer, Jeniffer Perez, and George Chenaux for performing some of the initial key experiments.
Disclosures The authors have no financial conflicts of interest.
This work was supported by National Institutes of Health Grants RO1-AR052755 and RO1-AI48644 and a Veterans Affairs merit grant.
Abbreviations used in this paper:
- bone marrow
- bone marrow-derived dendritic cell
- dendritic cell
- draining lymph node
- Flt3 ligand
- keyhole limpet hemocyanin
- Langerhans cell
- leukotriene C4
- parasite burden
- pyrrolidine dithiocarbamate
- recombinant murine
- RNase protection assay
- Received October 21, 2008.
- Accepted March 8, 2010.
- Copyright © 2010 by The American Association of Immunologists, Inc.